Automated decision making and profiling
Understanding GDPR
Shalini Kurapati, CIPP/E
Co-founder and CEO, Clearbox AI
AI: Lawfulness, fairness, and transparency
Legal basis
Fairness and discrimination - to infer data about people
Transparency
Special provisions- article 22
Article 22
Fairness and anti-discrimination obligations
Article 22 - automated individual decision-making, including profiling
Right to human intervention and explanation
1
Image source: flaticon.com
Profiling
Analyze and predict behavior, large-scale processing, using AI/machine learning
Identify and link behavior and attributes
Create profiles, and predict behavior based on profiles
Examples:
Treatment success
Risk of re-offending
1
Information Commissioner's Office
Ethical concerns related to profiling
Propagate or reinforce bias
Design, data, their inherent complexity/ black box
Gender, racial discrimination
1
Image source: bbc.com, propublica.com, and techcrunch.com
The famous profiling example
320,000 users took an online personality test for 'academic research'
Collected personal data of friends of the user, up to 50m user data without consent
Personality/political profiling, targeted ads to those likely to change mind
Used in many elections in the US, UK, and worldwide
1
Image source: The Guardian
Automated decision making
Decisions by automated means without human involvement
May or may not involve profiling
Any type of data - survey, location, profiling
Example
: Automated CV screening
1
ARTICLE 29 WORKING PARTY Guidelines on Automated individual decision-making and Profiling for the purposes of Regulation 2016/679
Banking example
Automated mortgage decision
Usually, AI gives yes or no answers, need for explanation and human intervention
Mechanisms for preventing bias and correcting and improving AI decisions
Let's practice!
Understanding GDPR
Preparing Video For Download...